
Hello Wisecrackers, Cannot resist shering with you my perspective. Computer is a logic device and logic builds complexity bottom up while graphics builds it top down. They are therefore antagonistic by nature and in the single case of their glorious interplay - The Euclidean Geometry - it blocked abstract thinking for two millennia (no zero, no negative numbers, no continuity). The implications of that facts are 3-fold: (1) Compositional GUI needs some non-geometric and perhaps non-commutative concepts (2) GUI will always set limit on the scope of your thinking (3) conceptually rich DSL is your best friend as human-computer interface. The role of GUI has gone way beyond its usefulness and now serves the industry to sell computers as candies. Apps like Acrobat Reader became little more than GUI when judged by size and only spymasters' conflicting interests strip some GUI excess to see you swim in sewers of the Internet-turned-video. This visual mania is dangerous as it discourages investigative thinking and reliance on concepts. It is shocking and in my judgement not a coincidence that major theories (like Chaos T.) were discovered using punched cards or plain pen and paper, while Artificial Intelligence research has been very acutely wounded by mouse. Down to Earth TTS reading should have human quality by now, had the researchers used conceptual and compositional approach. If linguists knew Haskell they could write DSLs to intelligently query large data sets and step by step discover human algorithms. Instead they use mostly statistics for pattern recognition. If experts could build DSLs for amplifying their own human concepts they would multiply computer power by human intelligence. Instead they are taught how to press GUI buttons. And what is behind this buttons? Statistics and designer crap! Down to the dust of your heels: make Acrobat Reader TTS read some PDF and you will hear main text intermingled with unrelated footnotes and paragraph titles and copyrights of every picture or graph repeated forever. You can correct this of course with several lines of Haskell code but then Wall Street should sell Adobe and buy You;-) I have been using Haskell for bizarre mathematical stuff but rarely feel the need for anything more than Unicode. Still I would appreciate greatly a simple compositional GUI like Clean has. Perhaps Haskell friendly drivers for selected cards is another path to simplicity? The ultimate GUI waits for operator algebra action where Haskell may show its Category Theoretical teeth. I doubt real world will ever learn Haskell so why Haskellers should pull every piece of real world crap? My advice is: stay clean, the world is wrong. Alleluia! Cheers, - Andrzej Jaworski

(3) conceptually rich DSL is your best friend as human-computer interface. I agree that a well-designed DSL is *my* best friend but I suspect it would be my grandma's worst enemy. And a badly-designed DSL is also everybody's worst enemy.
Also a traditional GUI is discoverable without a user-manual. I can open up MS Word for the first time and get around - Emacs, while vastly more powerful, is also vastly more opaque. -deech

aditya siram wrote:
And a badly-designed DSL is also everybody's worst enemy. Also a traditional GUI is discoverable without a user-manual.
I am glad you mentioned that. Perhaps I did not emphasised enough that DSL shold be user-specific knoledge discovery (bidirectionally)though transferable as skill. GUI excells in repetitive tasks but once you try something new it is a drag. I know that computers are and will mostly be used for repetitive tasks but they are becoming gadgets in these roles. -Andrzej

DSL
shold be user-specific knoledge discovery (bidirectionally)though transferable as skill. This is interesting. I am always looking for new ideas in GUI design. Could you elaborate on this? Do you have examples of where this has been successful? -deech

This is interesting. I am always looking for new ideas in GUI design. Could you elaborate on this? Do you have examples of where this has been successful?
Successful: quantitative finance and trading. Productive: unifying and scaling up from simple tools -Andrzej
participants (2)
-
aditya siram
-
Andrzej Jaworski